THUDM / GCC

GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training @ KDD 2020

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Regarding the choice of subgraph enhancement method.

chen1287 opened this issue · comments

Hello, I am trying to reproduce your code. I want to change the way of subgraph enhancement to neighbor sampling. However, when I changed the parameter of 'aug' in graph_dataset to 'ns', the pre-training process got stuck and the memory usage was abnormally high.